Part IV: Complete Solutions, Chapter 9 486 Chapter 9: Hypothesis Testing Section 9.1 Note: For all graphs provided, the P value is indicated by the shaded portion in the tails. 1. See text for definitions. Essays may include (a) A working hypothesis about the population parameter in question is called the null hypothesis. The value specified in the null hypothesis is often a historical value, a claim, or a production specification. (b) Any hypothesis that differs from the null hypothesis is called an alternate hypothesis. (c) If we reject the null hypothesis when it is in fact true, we have an error that is called a type I error. On the other hand, if we fail to reject the null hypothesis when it is in fact false, we have made an error that is called a type II error. (d) The probability with which we are willing to risk a type I error is called the level of significance of a test. The probability of making a type II error is denoted by . 2. The alternate hypothesis is used to determine which type of test is used. 3. No, if we fail to reject the null hypothesis, we have not proven it to be true beyond all doubt. The evidence is not sufficient to merit rejecting H 0 . 4. No, if we reject the null hypothesis, we have not proven it to be false beyond all doubt. The test was conducted with a level of significance that is the probability with which we are willing to risk a type I error (rejecting H 0 when it is in fact true). 5. (a) The claim is = 60 kg, so you would use H 0 : = 60 kg. (b) We want to know if the average weight is less than 60 kg, so you would use H1: < 60 kg. (c) We want to know if the average weight is greater than 60 kg, so you would use H1: > 60 kg. (d) We want to know if the average weight is different from 60 kg, so you would use H1: 60 kg. (e) Since part (b) is a left-tailed test, the critical region is on the left. Since part (c) is a right-tailed test, the critical region is on the right. Since part (d) is a two-tailed test, the critical region is on both sides of the mean. 6. (a) The claim is = 8.3 min, so you would use H 0 : = 8.3 min. If you believe that the average is less than 8.3 min, then you would use H1: < 8.3 min. This is a left-tailed test. (b) The claim is = 8.3 min, so you would use H 0 : = 8.3 min. If you believe the average is different from 8.3 min, then you would use H1: 8.3 min. This is a two-tailed test. (c) The claim is = 4.5 min, so you would use H 0 : = 4.5 min. If you believe the average is more than 4.5 min, then you would use H1: > 4.5 min. This is a right-tailed test. (d) The claim is = 4.5 min, so you would use H 0 : = 4.5 min. If you believe the average is different from 4.5 min, then you would use H1: 4.5 min. This is a two-tailed test. 7. (a) The claim is = 16.4 ft, so H 0 : = 16.4 ft. (b) You want to know if the average is getting larger, so H1: > 16.4 ft. (c) You want to know if the average is getting smaller, so H1: < 16.4 ft. (d) You want to know if the average is different from 16.4 ft, so H1: 16.4 ft. (e) Since part (b) is a right-tailed test, the area corresponding to the P value is on the right. Since part (c) is a left-tailed test, the area corresponding to the P value is on the left. Since part (d) is a two-tailed test, the area corresponding to the P value is on both sides of the mean. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 8. 487 (a) The claim is = 8.7 s, so H 0 : = 8.7 s. (b) You want to know if the average is longer, so H1: > 8.7 s. (c) You want to know if the average is reduced, so H1: < 8.7 s. (d) Since part (b) is a right-tailed test, the P value area is on the right. Since part (c) is a left-tailed test, the P value area is on the left. 9. (a) 0.01 H 0 : 4.7% H1: 4.7% Since > is in H1 , use a right-tailed test. (b) Use the standard normal distribution. We assume x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x n 5.38 4.7 2.4 10 0.90 (c) P value = P(z > 0.90) = 0.1841 (d) Since a P value of 0.1841 > 0.01 for , we fail to reject H 0 . The data are not statistically significant. (e) There is insufficient evidence at the 0.01 level to reject the claim that average yield for bank stocks equals average yield for all stocks. 10. (a) 0.05 H 0 : 85 mg/100 mL H1: 85 mg/100 mL Since > is in H1 , use a right-tailed test. (b) Use the standard normal distribution. We assume that x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x n 93.8 85 12.5 8 1.99 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 488 (c) P value = P(z > 1.99) = 0.0233 (d) Since a P value of 0.0233 0.05, reject H 0 . Yes, the data are statistically significant. (e) The sample evidence is sufficient at the 0.05 level to justify rejecting H 0 . It seems Gentle Ben’s glucose is higher than average. 11. (a) 0.01 H 0 : 4.55 g H1: 4.55 g Since < is in H1 , use a left-tailed test. (b) Use the standard normal distribution. We assume that x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x n 3.75 4.55 0.70 6 2.80 (c) P value = P(z < 2.80) = 0.0026 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 489 (d) Since a P value of 0.0026 0.01, we reject H 0 . Yes, the data are statistically significant. (e) The sample evidence is sufficient at the 0.01 level to justify rejecting H 0 . It seems the humming birds in the Grand Canyon region have a lower average weight. 12. (a) 0.05 H 0 : 19 H1: 19 Since < is in H1 , use a left-tailed test. (b) Use the standard normal distribution. We assume that x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x n 17.1 19 4.5 14 1.58 (c) P value = P(z < 1.58) = 0.0571 (d) Since a P value of 0.0571 > 0.05 for , we fail to reject H 0 . The data are not statistically significant. (e) There is insufficient evidence at the 0.05 level to reject H 0 . It seems the average P/E for large banks is not less than that of the S&P 500 Index. 13. (a) 0.01 H 0 : 11% H1: 11% Since is in H1 , use a two-tailed test. (b) Use the standard normal distribution. We assume that x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x n 12.5 11 5.0 16 1.20 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 490 (c) P value = 2P(z > 1.20) = 2(0.1151) = 0.2302 (d) Since a P value of 0.2302 > 0.01 for , we fail to reject H 0 . The data are not statistically significant. (e) There is insufficient evidence at the 0.01 level of significance to reject H 0 . It seems the average hail damage to wheat crops in Weld County matches the national average. 14. (a) 0.01 H 0 : 28 mL/kg H1: 28 mL/kg Since is in H1 , use a two-tailed test. (b) Use the standard normal distribution. We assume that x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x n 32.7 28 4.75 7 2.62 (c) P value = 2P(z > 2.62) = 2(0.0044) = 0.0088 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 491 (d) Since a P value of 0.0088 0.01, we reject H 0 . The data are statistically significant. (e) At the 1% level of significance, the sample average is sufficiently different from = 28 that we reject H 0 . It seems Roger’s average red cell volume is different from the average for healthy adults. Section 9.2 1. The P value for a two-tailed test of μ is twice the P value for a one-tailed test based on the same sample data and null hypothesis. 2. If σ is known, use the standard normal distribution. If σ is not known, use the Student’s t distribution with n – 1 degrees of freedom. 3. Use n – 1 degrees of freedom. 4. Not necessarily. If the P value = 0.04, you would reject at the α = 0.05 level but not at the α = 0.01 level. On the other hand, if the P value is 0.003, you would reject at both the α = 0.05 and α = 0.01 levels. 5. Yes. If the P value is less than α = 0.01, then it also will be less than α = 0.05. In both cases, reject the null hypothesis. 6. No. The P value for the two-tailed test is twice the P value for the one-tailed test. If the P value < 0.01, then 2 × P value might be greater than or less than 0.01. 7. (a) 0.01 H 0 : 16.4 ft H1: 16.4 ft (b) Use the standard normal distribution. The sample size is large, n 30, and = 3.5. x 17.3 16.4 z 1.54 n 3.5 36 (c) P value = P(z > 1.54) = 0.0618 (d) Since a P value of 0.0618 > 0.01, we fail to reject H 0 . The data are not statistically significant. (e) At the 1% level of significance, there is insufficient evidence to say average storm level is increasing. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 492 8. (a) 0.01 H 0 : 38 h H1: 38 h (b) Use the standard normal distribution. The sample size is large, n 30, and σ = 1.2 hours. x 37.5 38 z 2.86 n 1.2 47 (c) P value = P(z < 2.86) = 0.0021 (d) Since a P value of 0.0021 0.01, we reject H 0 . The data are statistically significant. (e) At the 1% level of significance, evidence shows average assembly time is less than 38 hours. 9. (a) 0.05 H 0 : 41.7 H1: 41.7 (b) Use a standard normal distribution. The sample size is large, n 30, and σ = 18.5 e-mails. x 36.2 41.7 z 1.99 n 18.5 45 (c) P value = 2P(z < 1.99) = 2(0.0233) = 0.0466 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 493 (d) Since a P value of 0.0466 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, there is sufficient evidence to say the average number of e-mails is different than 41.7 under the new priority system. 10. (a) 0.05 H 0 : 7.4 pH H1: 7.4 pH (b) Use the Student’s t distribution with d.f. = n 1 = 31 1 = 30. The sample size is large, n 30, and is unknown. x 8.1 7.4 t 2.051 s n 1.9 31 (c) If d.f. = 30, 2.051 falls between the entries 2.042 and 2.457, use two-tailed areas to find that 0.020 < P value < 0.050. Using a TI-84, P value ≈ 0.0491. (d) Since the entire P-value interval 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to say that the drug has changed the mean pH level. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 494 11. (a) 0.01 H 0 : 1.75 years H1: 1.75 years (b) Use the Student’s t distribution with d.f. = n 1 = 46 1 = 45. The sample size is large, n 30, and is unknown. x 2.05 1.75 t 2.481 s n 0.82 46 (c) For d.f. = 45, 2.481 falls between entries 2.412 and 2.690. Use one-tailed areas to find that 0.005 < P value < 0.010. Using a TI-84, P value ≈ 0.0084. (d) Since the entire P-value interval 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, sample data indicate that the average age of Minnesota region coyotes is greater than 1.75 years. 12. (a) 0.05 H 0 : 19 in. H1: 19 in. (b) Use the Student’s t distribution with d.f. = n 1 = 51 1 = 50. The sample size is large, n 30, and is unknown. x 18.5 19 t 1.116 s n 3.2 51 (c) For d.f. = 50, 1.116 falls between entries 0.679 and 1.164. Use one-tail areas to find that 0.125 < P value < 0.250. Using a TI-84, P value ≈ 0.1349. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 495 (d) Since the P-value interval > 0.05, we fail to reject H 0 . The data are not statistically significant. (e) At the 5% level of significance, the sample data do not indicate that the average fish length is less than 19 inches. 13. (a) 0.05 H 0 : 19.4 H1: 19.4 (b) Use the Student’s t distribution with d.f. = n 1 = 36 1 = 35. The sample size is large, n 30, and is unknown. x 17.9 19.4 t 1.731 s n 5.2 36 (c) For d.f. = 35, 1.731 falls between entries 1.690 and 2.030. Use two-tailed areas to find that 0.05 < P value < 0.100. Using a TI-84, P value ≈ 0.0923. (d) Since the P-value interval > 0.05, we fail to reject H 0 . The data are not statistically significant. (e) At the 5% level of significance, the sample evidence does not support rejecting the claim that the average P/E for socially responsible funds is different from that of the S&P stock index. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 496 14. (a) 0.01 H 0 : 8.0 ppb H1: 8.0 ppb (b) Use the Student’s t distribution with d.f. = n 1 = 37 1 = 36. The sample size is large, n 30, and is unknown. x 7.2 8.0 t 2.561 s n 1.9 37 (c) In Table 6, use the closest d.f. smaller than 36 or d.f. = 35. Since 2.561 falls between entries 2.438 and 2.724, use one-tailed areas to find 0.005 < P value < 0.010. Using a TI-84, P value ≈ 0.0074. (d) Since the entire P-value interval 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, sample data support the claim that the average arsenic content is less than 8.0 ppb. 15. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.05 H 0 : 4.8 H1: 4.8 (b) Use the Student’s t distribution with d.f. = n 1 = 6 1 = 5. We assume that x has a distribution that is approximately normal and that is unknown. x 4.40 4.8 t 3.499 s n 0.28 6 (c) For d.f. = 5, 3.499 falls between entries 3.365 and 4.032. Use one-tailed areas to find that 0.005 < P value < 0.010. Using a TI-84, P value ≈ 0.0086. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 497 Student's t Distribution with d.f. = 5 -3.499 -2.000 -1.000 0.000 1.000 2.000 3.500 (d) Since the entire P-value interval 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, sample evidence supports the claim that the average RBC count for this patient is less than 4.8. 16. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.01 H 0 : 14 H1: 14 (b) Use the Student’s t distribution with d.f. = n 1 = 10 1 = 9. We assume that x has a distribution that is approximately normal and that is unknown. x 15.1 14 t 1.386 s n 2.51 10 (c) For d.f. = 9, 1.386 falls between entries 1.383 and 1.574. Use one-tailed areas to find that 0.075 < P value < 0.100. Using a TI-84, P value ≈ 0.0996. (d) Since P-value interval > 0.01, we fail to reject H 0 . No, the data are not statistically significant. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 498 (e) At the 1% level of significance, the sample data do not support the claim that the average HC for this patient is higher than 14. 17. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.01 H 0 : 67 H1: 67 (b) Use the Student’s t distribution with d.f. = n 1 = 16 1 = 15. We assume that x has a distribution that is approximately normal and that is unknown. x 61.8 67 t 1.962 s n 10.6 16 (c) For d.f. = 15, 1.962 falls between entries 1.753 and 2.131. Use two-tailed areas to find that 0.050 < P value < 0.100. Using a TI-84, P value ≈ 0.0686. (d) Since P-value interval > 0.01, we fail to reject H 0 . The data are not statistically significant. (e) At the 1% level of significance, sample evidence does not support a claim that the average thickness of slab avalanches in Vail is different from that of those in Canada. 18. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.05 H 0 : 77 years H1: 77 years (b) Use the Student’s t distribution with d.f. = n 1 = 20 1 = 19. We assume that x has a distribution that is approximately normal and that is unknown. x 71.4 77 t 1.213 s n 20.65 20 (c) For d.f. = 19, 1.213 falls between entries 1.187 and 1.328. Use one-tailed areas to find that 0.100 < P value < 0.125. Using a TI-84, P value ≈ 0.1200. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 499 (d) Since the P-value interval > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, evidence is not strong enough to conclude that the population mean life span is less than 77 years. 19. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.05 H 0 : 8.8 H1: 8.8 (b) Use the Student’s t distribution with d.f. = n 1 = 14 1 = 13. We assume that x has a distribution that is approximately normal and that is unknown. x 7.36 8.8 t 1.337 s n 4.03 14 (c) For d.f. = 13, 1.337 falls between entries 1.204 and 1.350. Use two-tailed areas to find that 0.200 < P value < 0.250. Using a TI-84, P value ≈ 0.2042. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 500 (d) Since P-value interval > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, we cannot conclude that the catch is different from 8.8 fish per day. 20. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.01 H 0 : 1,300 H1: 1,300 (b) Use the Student’s t distribution with d.f. = n 1 = 10 1 = 9. We assume that x has a distribution that is approximately normal and that is unknown. x 1268 1300 t 2.714 s n 37.29 10 (c) For d.f. = 9, 2.714 falls between entries 2.262 and 2.821. Use two-tailed areas to find 0.020 < P value < 0.050. Using a TI-84, P value ≈ 0.0239. (d) Since P-value interval > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, there is not enough evidence to conclude that the population mean of tree ring dates is different from 1,300. 21. (a) The P value of a one-tailed test is smaller. For a two-tailed test, the P value is double because it includes the area in both tails. (b) Yes; the P value of a one-tailed test is smaller, so it might be smaller than , whereas the P value of a two-tailed test is larger than . (c) Yes; if the two-tailed P value is less than , the one-tail area is also less than . (d) Yes, the conclusions can be different. The conclusion based on the two-tailed test is more conservative in the sense that the sample data must be more extreme (differ more from H 0 ) in order to reject H 0 . 22. Essay or class discussion. 23. (a) H 0 : 20 H1: 20 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 501 For = 0.01, c = 1 0.01 = 0.99, 4, and zc 2.58. E zc n 2.58 4 36 1.72 xExE 22 1.72 22 1.72 20.28 23.72 The hypothesized mean = 20 is not in the interval. Therefore, we reject H 0 . (b) Because n = 36 is large, the sampling distribution of x is approximately normal by the central limit theorem, and we know σ. x 22 20 z 3.00 / n 4/ 36 From Table 5, P value = 2P(z < 3.00) = 2(0.0013) = 0.0026. Since 0.0026 0.01, we reject H 0 . The results are the same. 24. (a) H 0 : 21 H1: 21 For = 0.01, c = 1 0.01 = 0.99, 4, and zc 2.58. E zc n 2.58 4 36 1.72 xExE 22 1.72 22 1.72 20.28 23.72 The hypothesized mean = 21 falls into the confidence interval. Therefore, we do not reject H 0 . (b) For = 0.01, the two-tailed test’s critical values are z0 2.58. Because n = 36 is large, the sampling distribution of x is approximately normal by the central limit theorem, and we know σ. x 22 21 z 1.50 / n 4/ 36 From Table 5, P value = 2P(z < 1.50) = 2(0.0668) = 0.1336. Since 0.1336 > 0.01, we do not reject H 0 . The results are the same. 25. For a right-tailed test and = 0.01, the critical value is z0 2.33; critical region is values to the right of 2.33. Since the sample statistic z = 1.54 is not in the critical region, fail to reject H 0 . At the 1% level, there is insufficient evidence to say that the average storm level is increasing. Conclusion is the same as with the P-value method. 26. For a left-tailed test and = 0.01, the critical value z0 2.33; critical region is values to the left of 2.33. Since the sample statistic z = 2.86 is in the critical region, reject H 0 . At the 1% level, evidence shows that average assembly time is less than 38 hours. The conclusion is that same as with the P-value method. 27. For a two-tailed test and = 0.05, the critical value z0 1.96; critical regions are values to the left of 1.96 and values to the right of 1.96. Since the sample test statistic z = 1.99 is in the critical region, reject H 0 . At the 5% level, there is sufficient evidence to say the average number of e-mails is different with the new priority system. The conclusion is the same as with the P-value method. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 502 28. For a two-tailed test and = 0.05, critical values are t0 2.042 with d.f. = 30; critical regions are values to the right of 2.042 and those to the left of 2.042. Since the sample test statistic t = 0.051 is in the critical region, reject H 0 . At the 5% level, the evidence is sufficient to say that the drug has changed the mean pH level. The conclusion is the the same as with the P-value method. 29. For a right-tailed test and = 0.01, critical value is t0 2.412 with d.f. = 45. Critical region is values to the right of 2.412. Since the sample test statistic t = 2.481 is in the critical region, reject H 0 . At the 1% level, sample data indicate that the average age of Minnesota region coyotes is higher than 1.75 years. The conclusion is the same as with the P-value method. 30. For a left-tailed test and = 0.05, critical value is t0 1.676 with d.f. = 50. Critical region is values to the left of 1.676. Since the sample test statistic t = 1.116 is not in the critical region, fail to reject H 0 . At the 5% level of significance, the sample data do not indicate that the average fish length is less than 19 inches. The conclusion is the same as with the P-value method. Section 9.3 1. 2. The value of p comes from H0. Note that q = 1 – p. pˆ r n 3. Yes. The corresponding P value for a one-tailed test is half that of a two-tailed test. Thus the P value for the one-tailed test is also less than 0.01. 4. Answers vary. First, we don’t know if the study is based on population data or sample data. If the study is based on population data, we do not need to conduct a hypothesis test. However, assuming that the study was based on sample data, testing H0: p = 0.15 versus H1: p > 0.15 seems appropriate. Without a specific source for the study, we do not know how reliable the information in the sample is. Also, we are not given information about the sample size. Level of significance could be one of the common values, α = 0.01 or α = 0.05. Finally, if the conclusions are based on sample data, we cannot conclude that the conclusions are absolutely true. 5. (i) (a) 0.01 H 0 : p 0.301 H1: p 0.301 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 215(0.301) 647 and nq = 215(0.699) 150.3 are both > 5. r 46 pˆ 0.214 n 215 pˆ p 0.214 0.301 z 2.78 pq n 0.301(0.699) 215 (c) P value = P(z < 2.78) = 0.0027 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 503 (d) Since a P value of 0.0027 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the sample data indicate that the population proportion of numbers with a leading 1 in the revenue file is less than 0.301 predicted by Benford’s law. (ii) Yes. The revenue data file seems to include more numbers with higher first nonzero digits than Benford’s law predicts. (iii) We have not proved H 0 to be false. However, because our sample data lead us to reject H 0 and conclude that there are too few numbers with a leading digit 1, more investigation is merited. 6. (i) (a) 0.01 H 0 : p 0.301 H1: p 0.301 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is, because np = 228(0.301) 68.6 and nq = 228(0.699) 159.4 are both > 5. r 92 pˆ 0.404 n 228 pˆ p 0.404 0.301 z 3.39 pq n 0.301(0.699) 228 (c) P value = P(z > 3.39) = 0.0003 Standard Normal Distribution -3.50 -3.00 -2.00 -1.00 0.00 Copyright © Houghton Mifflin Company. All rights reserved. 1.00 2.00 3.00 3.39 Part IV: Complete Solutions, Chapter 9 504 (d) Since a P value of 0.0003 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the sample data indicate that the proportion of numbers in the revenue file with a leading digit 1 exceed the 0.301 predicted by Benford’s law. (ii) Yes. There seem to be too many entries with a leading digit 1. (iii) We have not proved H 0 to be false. However, because our data led us to reject H 0 and conclude that there are “too many” numbers with a leading digit 1, more investigation is merited. 7. (a) 0.01 H 0 : p 0.70 H1: p 0.70 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 32(0.7) = 22.4 and nq = 32(0.3) = 9.6, and both are greater than 5. r 24 pˆ 0.75 n 32 pˆ p 0.75 0.70 z 0.62 pq n 0.70(0.30) 32 (c) P value = 2P(z > 0.62) = 2(0.2676) = 0.5352 (d) Since a P value of 0.5352 > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, we cannot say that the population proportion of arrests of males aged 15 to 34 in Rock Springs is different than 70%. 8. (a) 0.05 H 0 : p 0.67 H1: p 0.67 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 38(0.67) = 25.46 and nq = 38(0.33) = 12.54, and both are greater than 5. r 21 pˆ 0.5526 n 38 pˆ p 0.5526 0.67 z 1.54 pq n 0.67(0.33) 38 (c) P value = P(z < 1.54) = 0.0618 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 505 (d) Since a P value of 0.0618 > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, there is insufficient evidence to say that the proportion of women athletes who graduate is less than 67%. 9. (a) 0.01 H 0 : p 0.77 H1: p 0.77 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 27(0.77) = 20.79 and nq = 27(0.23) = 6.21, and both are greater than 5. r 15 pˆ 0.5556 n 27 pˆ p 0.5556 0.77 z 2.65 pq n 0.77(0.23) 27 (c) P value = P(z < 2.65) = 0.0004 (d) Since a P value of 0.0004 0.01, we reject H 0 . Yes, the data are statistically significant. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 506 (e) At the 1% level of significance, the data show that the population proportion of driver fatalities related to alcohol is less than 77% in Kit Carson County. 10. (a) 0.05 H 0 : p 0.24 H1: p 0.24 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 56(0.24) = 13.44 and nq = 56(0.76) = 42.56, and both are greater than 5. r 12 pˆ 0.2143 n 56 pˆ p 0.2143 0.24 z 0.45 pq n 0.24(0.76) 56 (c) P value = 2P(z < 0.45) = 2(0.3264) = 0.6528 (d) Since a P value of 0.6528 > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the data do not indicate that the proportion of college students favoring the color blue is different from 0.24. 11. (a) 0.01 H 0 : p 0.50 H1: p 0.50 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 34(0.50) = 17 and nq = 17, and both are greater than 5. r 10 pˆ 0.2941 n 34 pˆ p 0.2941 0.50 z 2.40 pq n 0.5(0.5) 34 (c) P value = P(z < 2.40) = 0.0082 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 507 (d) Since a P value of 0.0082 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the data indicate that the population proportion of female wolves is now less than 50% in the region. 12. (a) 0.05 H 0 : p 0.75 H1: p 0.75 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 83(0.75) = 62.25 and nq = 83(0.25) = 20.75, and both are greater than 5. r 64 pˆ 0.7711 n 83 pˆ p 0.7711 0.75 z 0.44 pq n 0.75(0.25) 83 (c) P value = 2P(z > 0.44) = 2(0.3300) = 0.6600 (d) Since a P value of 0.6600 > 0.05, we fail to reject H 0 . No, the data are not statistically different. (e) At the 5% level of significance, there is insufficient evidence to indicate that the population proportion of guests who catch pike of 20 pounds is different from 75%. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 508 13. (a) 0.01 H 0 : p 0.261 H1: p 0.261 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 317(0.261) = 82.737 and nq = 317(0.739) = 234.263, and both are greater than 5. r 61 pˆ 0.1924 n 317 pˆ p 0.1924 0.261 z 2.78 pq n 0.261(0.739) 317 (c) P value = 2P(z < 2.78) = 2(0.0027) = 0.0054 (d) Since a P value of 0.0054 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the sample data indicate that the population proportion of the fivesyllable sequence is different from the text of Plato’s Republic. 14. (a) 0.01 H 0 : p 0.214 H1: p 0.214 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 493(0.214) = 105.502 and nq = 493(0.786) = 387.498, and both are greater than 5. r 136 pˆ 0.2759 n 493 pˆ p 0.2759 0.214 z 3.35 pq n 0.214(0.786) 493 (c) P value = P(z > 3.35) = 0.0004 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 509 Standard Normal Distribution -4.00 -3.00 -2.00 -1.00 0.00 1.00 2.00 3.35 4.00 (d) Since a P value of 0.0004 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the sample data indicate that the population proportion of the fivesyllable sequence is higher than that found in Plato’s Symposium. 15. (a) 0.01 H 0 : p 0.47 H1: p 0.47 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 1006(0.47) = 472.82 and nq = 1006(0.53) = 533.18, and both are greater than 5. r 490 pˆ 0.4871 n 1006 pˆ p 0.4871 0.47 z 1.09 pq n 0.47(0.53) 1006 (c) P value = P(z > 1.09) = 0.1379 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 510 (d) Since a P of 0.1379 > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, there is insufficient evidence to support the claim that the population proportion of customers loyal to Chevrolet is more than 47%. 16. (a) 0.05 H 0 : p 0.80 H1: p 0.80 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 115(0.8) = 92 and nq = 115(0.2) = 23, and both are greater than 5. r 88 pˆ 0.7652 n 115 pˆ p 0.7652 0.80 z 0.93 pq n 0.80(0.20) 115 (c) P value = P(z < 0.93) = 0.1762 (d) Since a P value of 0.1762 > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, there is insufficient evidence that the population proportion of prices ending with digits 9 or 5 is less than 80%. 17. (a) 0.05 H 0 : p 0.092 H1: p 0.092 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 196(0.092) = 18.032 and nq = 196(0.908) = 177.968, and both are greater than 5. r 29 pˆ 0.1480 n 196 pˆ p 0.1480 0.092 z 2.71 pq n 0.092(0.908) 196 (c) P value = P(z > 2.71) = 0.0034 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 511 (d) Since a P value of 0.0034 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the data indicate that the population proportion of students with hypertension during final exams week is higher than 9.2%. 18. (a) 0.01 H 0 : p 0.12 H1: p 0.12 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 209(0.12) = 25.08 and nq = 209(0.88) = 183.92, and both are greater than 5. r 16 pˆ 0.0766 n 209 pˆ p 0.0766 0.12 z 1.93 pq n 0.12(0.88) 209 (c) P value = P(z < 1.93) = 0.0268 (d) Since a P value of 0.0268 > 0.01, we fail to reject H 0 . No, the data are not statistically significant. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 512 (e) At the 1% level of significance, the data are insufficient to conclude that the population proportion of patients having headaches is less than 0.12. 19. (a) 0.01 H 0 : p 0.82 H1: p 0.82 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 73(0.82) = 59.86 and nq = 73(0.18) = 13.14, and both are greater than 5. r 56 pˆ 0.7671 n 73 pˆ p 0.7671 0.82 z 1.18 pq n 0.82(0.18) 73 (c) P value = 2P(z 1.18) = 2(0.1190) = 0.2380 (d) Since a P value of 0.2380 > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, the evidence is insufficient to indicate that the population proportion of extroverts among college student government leaders is different from 82%. 20. For a two-tailed test and = 0.01, critical values are z0 2.58. The critical regions are values greater than 2.58 and values less than 2.58. The sample test statistic z = 0.62 is not in the critical region, so we do not reject H 0 . Result is consistent with the P-value conclusion. 21. For a left-tailed test and = 0.01, critical value is z0 2.33. The critical region consists of values less than 2.33. The sample test statistic z = 2.65 is in the critical region, so we reject H 0 . Result is consistent with the P-value conclusion. 22. For a two-tailed test and = 0.01, critical value is z0 2.33. The critical region consists of values greater than 2.33. The sample test statistic z = 1.09 is not in the critical region, so we fail to reject H 0 . Result is consistent with the P-value conclusion. Section 9.4 1. Paired data are dependent. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 513 2. Take the difference of corresponding paired data values. The sample test statistics is d , which is the mean of the differences. 3. H0: d 0 . We test that the mean difference is 0. 4. Here, n is the number of data pairs. 5. Here, d.f. = n – 1. 6. (a) For a right-tailed test that the before score is higher, use d = B – A. (b) For a left-tailed test that the before score is higher, use d = A – B. 7. (a) 0.05 H 0 : d 0 H1: d 0 Since is in H1 , a two-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d=BA 1 3 2 2 3 5 4 4 5 10 6 15 7 6 8 7 d 2.25, sd 7.78 t d d 2.25 0 0.818 sd n 7.78 8 (c) d.f. = n 1 = 8 1 = 7 From Table 6 in Appendix II, 0.818 falls between entries 0.711 and 1.254. Using two-tailed areas, find that 0.250 < P value < 0.500. Using a TI-84, P value ≈ 0.4402. (d) Since the P-value interval is > 0.05, we fail to reject H 0 . No, the data are not statistically significant. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 514 (e) At the 5% level of significance, the evidence is insufficient to claim a difference in population mean percentage increases for corporate revenue and CEO salary. 8. (a) 0.01 H 0 : d 0 H1: d 0 Since is in H1 , a two-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d=BA 1 0.1 2 0.4 3 0.4 4 1.0 5 0.6 6 0.6 7 0.5 d 0.37, sd 0.47 t d d sd n 0.37 0 0.47 7 2.08 (c) d.f. = n 1 = 7 1 = 6 From Table 6 in Appendix II, 2.08 falls between entries 1.943 and 2.447. Using two-tailed areas, find that 0.050 < P value < 0.100. Using a TI-84, the P value ≈ 0.0823. (d) Since the P-value interval is > 0.01, we do not reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, the evidence is insufficient to claim that there is a difference in population mean hours per fish between boat fishing and shore fishing. 9. (a) 0.01 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d = Jan April 1 35 2 9 3 26 4 24 5 17 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 515 d 12.6, sd 22.66 t d d 12.6 0 1.243 sd 22.66 5 n (c) d.f. = n 1 = 5 1 = 4 From Table 6 in Appendix II, 1.243 falls between entries 0.741 and 1.344. Using one-tailed areas, find that 0.125 < P value < 0.250. Using a TI-84, the P value ≈ 0.1408. (d) Since the P-value interval is > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, the evidence is insufficient to claim average peak wind gusts are higher in January. 10. (a) 0.05 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d=BA 1 1.2 2 1.2 3 2.9 4 1.7 5 13.4 6 2.8 7 5.3 8 7.9 9 6.7 10 4.3 d 4.50, sd 4.1226 t d d sd n 4.50 0 4.1226 10 3.452 (c) d.f. = n 1 = 10 1 = 9 From Table 6 in Appendix II, 3.452 falls between entries 3.250 and 4.781. Using two-tailed areas, find that 0.0005 < P value < 0.005. Using a TI-84, P value ≈ 0.0036. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 516 Student's t Distribution, d.f. = 9 -4.000 -3.000 -2.000 -1.000 0.000 1.000 2.000 3.452 4.000 (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to claim that the January mean population of deer has dropped. Note that this test does not determine the cause of the drop. Development around the highway, disease, etc. may have affected the deer population. 11. (a) 0.05 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d = winter summer 1 19 2 4 3 17 4 7 5 9 6 0 7 9 8 10 d 6.125, sd 9.83 t d d 6.125 0 1.76 sd n 9.83 8 (c) d.f. = n 1 = 8 1 = 7 From Table 6 in Appendix II, 1.76 falls to the between entries 1.617 and 1.895. Using one-tailed areas, find that 0.05 < P value < 0.075. Using a TI-84, P value ≈ 0.0607. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 517 (d) Since the P-value interval is > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to indicate that the population average percentage of male wolves in winter is higher. 12. (a) 0.01 H 0 : d 0 H1: d 0 Since is in H1 , a two-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d=AB 1 2.9 2 1.1 3 2.1 10 1.3 4 2.1 11 4.0 5 1.4 12 5.6 6 1.8 13 7.6 7 3.3 8 5.1 14 0.9 9 1.6 15 4.1 16 6.5 d 1.10, sd 3.745 t d d 1.10 0 1.175 sd n 3.745 16 (c) d.f. = n 1 = 16 1 = 15 From Table 6 in Appendix II, 1.175 falls between entries 0.691 and 1.197. Use two-tailed areas to find that 0.250 < P value < 0.500. Using a TI-84, P value ≈ 0.2584. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 518 (d) Since the P-value interval is > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, the evidence is insufficient to claim that the population average birth and death rates are different in this region. 13. (a) 0.05 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d = houses hogans 1 5 2 2 3 22 4 23 5 4 6 19 7 33 8 32 d 6, sd 21.5 t d d 6 0 0.789 sd n 21.5 8 (c) d.f. = n 1 = 8 1 = 7 From Table 6 in Appendix II, 0.789 falls between entries 0.711 and 1.254. Use one-tailed areas to find that 0.125 < P value < 0.250. Using a TI-84, P value ≈ 0.2282. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 519 (d) Since the P-value interval is > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to show that the mean number of inhabited houses is greater than that of hogans. 14. (a) 0.05 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d = flaked nonflaked 1 4 2 1 3 9 4 2 5 3 6 6 7 21 8 13 d 6.125, sd 8.079 t d d 6.125 0 2.144 sd n 8.079 8 (c) d.f. = n 1 = 8 1 = 7 From Table 6 in Appendix II, 2.144 falls between entries 1.895 and 2.365. Use one-tailed areas to find that 0.025 < P value < 0.05. Using a TI-84, P value ≈ 0.0346. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 520 (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to indicate that the population average number of flaked stone tools is higher. 15. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.05 H0: 0 H1: 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. The sample size is greater than 30. d 2.472, sd 12.124 t d d 2.472 0 1.223 sd n 12.124 36 (c) d.f. = n 1 = 36 1 = 35 From Table 6 in Appendix II, 1.223 falls between entries 1.170 and 1.306. Use one-tailed areas to find that 0.100 < P value < 0.125. Using a TI-84, P value ≈ 0.1147. (d) Since the P-value interval > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to claim that the population mean cost of living index for housing is higher than that for groceries. 16. (i) Use a calculator to verify. Rounded values are used in part (ii). (ii) (a) 0.05 H0: 0 H1: 0 Since < is in H1 , a left-tailed test is used. (b) Use the Student’s t distribution. The sample size is greater than 30. d 5.7391, sd 15.910 t d d 5.7391 0 2.447 sd n 15.910 46 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 521 (c) d.f. = 45 From Table 6 in Appendix II, 2.447 falls between entries 2.412 and 2.690. Use one-tailed areas to find that 0.005 < P value < 0.010. Using a TI-84, P value ≈ 0.0092. (d) Since the P-value interval 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to claim that the population mean cost of living index for utilities is less than that for transportation in this region. 17. (a) 0.05 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d=BA 1 7 2 2 3 9 4 0 5 6 6 1 7 3 8 3 9 3 d 2.0, sd 4.5 t d d sd n 2.0 0 4.5 9 1.33 (c) d.f. = n 1 = 9 1 = 8 From Table 6 in Appendix II, 1.33 falls between entries 1.240 and 1.397. Use one-tailed areas to find that 0.100 < P value < 0.125. Using a TI-84, P value ≈ 0.1096. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 522 (d) Since the P-value interval is > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to claim that the population score on the last round is higher than that on the first. 18. (a) 0.05 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d=15 1 0.6 2 0.5 3 0.1 4 0.2 5 0.7 6 0.9 d 0.4, sd 0.447 t d d sd n 0.4 0 0.447 6 2.192 (c) d.f. = n 1 = 6 1 = 5 From Table 6 in Appendix II, 2.192 falls between entries 2.015 and 2.571. Use one-tailed areas to find that 0.025 < P value < 0.050. Using a TI-84, P value ≈ 0.0400. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 523 (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to claim that the population mean time for rats receiving larger rewards to run the maze is less. 19. (a) 0.05 H 0 : d 0 H1: d 0 Since > is in H1 , a right-tailed test is used. (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d = time 1 time 5 1 1.4 2 1.7 3 0.8 4 1.5 5 0.5 6 0.1 7 1.7 8 1.3 d 0.775, sd 1.0539 t d d 0.775 0 2.080 sd n 1.0539 8 (c) d.f. = n 1 = 8 1 = 7 From Table 6 in Appendix II, 2.080 falls between entries 1.895 and 2.365. Use one-tailed areas to find 0.025 < P value < 0.050. Using a TI-84, P value ≈ 0.0380. (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to claim that the population mean time for rats receiving larger rewards to climb the ladder is less. 20. (a) d 2.25; sd 7.778 ; t0.95 = 2.365 with d.f. = 7; E 2.365; 95% confidence interval for d from –4.25 to 8.75. We are 95% confident that the difference in population mean percentage increase between company revenue and CEO salaries is between –4.25% and 8.75%. (b) = 0.05; c = 1 – = 0.95; H0: d = 0; H1: d ≠ 0; Since d = 0 from the null hypothesis is in the 95% confidence interval, do not reject H0 at the 5% level of significance. The data do not indicate a difference in population mean percentage increases between company revenue and CEO salaries. This result is consistent with the conclusion reached using the P-value method of testing for = 0.05 in Problem 7. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 524 21. For a two-tailed test with = 0.05 and d.f. = 7, critical values are t0 2.365. The sample test statistic t = 0.818 is between 2.365 and 2.365, so we do not reject H 0 . This conclusion is the same as that reached by the P-value method. 22. For a right-tailed test with = 0.01 and d.f. = 4, the critical value is t0 3.747. The sample test statistic t = 1.243 is to the left of t0 , so we do not reject H 0 . This conclusion is the same as that reached by the P-value method. Section 9.5 1. The test statistic is x1 x2 . 2. Use the Student’s t distribution. 3. H0: μ1 = μ2 or H0: μ1 – μ2 = 0. 4. r r The test statistic is pˆ1 pˆ 2 , where pˆ1 1 and pˆ 2 2 . n1 n2 5. r r The best estimate is p 1 2 . n1 n2 6. (a) Satterthwaite’s gives the larger degrees of freedom. It is more conservative to use a smaller degrees of freedom because there will be more area in the tails of the Student’s t distribution, resulting in larger P values. (b) The P value will be larger. Using a larger P value might mean that the test is not significant, whereas a smaller P value might result in significance. 7. H1: μ1 > μ2 or H1: μ1 – μ2 > 0. 8. H1: μ1 < μ2 or H1: μ1 – μ2 < 0. 9. (a) 0.01 H 0 : 1 2 H1: 1 2 (b) Use the standard normal distribution. We assume that both population distributions are approximately normal and that 1 and 2 are known. x1 x2 2.8 2.1 0.7 z ( x1 x2 ) ( 1 2 ) 12 22 0.7 0 0.52 10 n 2 (c) P value = P(z > 2.57) = 0.0051 n1 2 2.57 0.7 10 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 525 (d) Since P value = 0.0051 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the evidence is sufficient to indicate that the population mean REM sleep time for children is more than for adults. 10. (a) 0.01 H 0 : 1 2 H1: 1 2 (b) Use the standard normal distribution. We assume that both population distributions are approximately normal and that 1 and 2 are known. x1 x2 43 36 7 z ( x1 x2 ) ( 1 2 ) 22 70 0.96 212 152 12 14 n1 n2 (c) P value = 2P(z > 0.96) = 2(0.1685) = 0.3370 12 (d) Since P value = 0.3370 > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, the evidence is insufficient to claim that there is a difference in mean population pollution index for Englewood and Denver. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 526 11. (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the standard normal distribution. We assume that both population distributions are approximately normal and that 1 and 2 are known. x1 x2 4.9 4.3 0.6 z ( x1 x2 ) ( 1 2 ) 12 22 0.6 0 1.52 46 2 2.16 1.2 51 n 2 (c) P value = 2P(z > 2.16) = 2(0.0154) = 0.0308 n1 (d) Since P value = 0.0308 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to show that there is a difference between mean response regarding preference for camping or fishing. 12. (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the standard normal distribution. We assume that both population distributions are approximately normal and that 1 and 2 are known. x1 x2 15.2 19.7 4.5 z ( x1 x2 ) ( 1 2 ) 12 n1 22 n 2 4.5 0 7.22 32 2 5.2 35 2.91 (c) P value = P(z < 2.91) = 0.0018 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 527 (d) Since P value = 0.0018 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to show that the population mean percentage of young adults who attend college is higher. 13. (i) Use a calculator to verify. Use rounded results to compute t. (ii) (a) 0.01 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume that both population distributions are approximately normal and that 1 and 2 are unknown. x1 x2 3.51 3.87 0.36 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n 2 0.36 0 0.812 10 0.94 12 2 0.965 (c) Since n1 n2 , d . f . n1 1 10 1 9 In Table 6 in Appendix II, 0.965 falls between entries 0.703 and 1.230. Use one-tailed areas to find that 0.125 < P value < 0.250. Using a TI-84, d.f. ≈ 19.96, and P value ≈ 0.1731. (d) Since the P-value interval > 0.01, we fail to reject H 0 . No, the data are not statistically significant. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 528 (e) At the 1% level of significance, the evidence is insufficient to indicate that violent crime in the Rocky Mountain region is higher than in New England. 14. (i) Use a calculator to verify. Use rounded results to compute t. (ii) (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume that both population distributions are approximately normal and that 1 and 2 are unknown. x1 x2 109.50 99.36 10.14 t ( x1 x2 ) ( 1 2 ) s12 s22 10.14 0 15.412 16 n n1 2 2 11.57 14 2.053 (c) Since n1 n2 , d . f . n1 1 14 1 13 . In Table 6 in Appendix II, 2.053 falls between entries 1.771 and 2.160. Use one-tailed areas find that 0.025 < P value < 0.050. Using a TI-84, d.f. ≈ 27.42, and P value ≈ 0.0249. to (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to indicate that the population mean rate of hay fever is lower for the age group over 50. 15. (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. Both sample sizes are large, ni 30, and 1 and 2 are unknown. x1 x2 344.5 354.2 9.7 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n 2 9.7 0 49.12 30 50.92 30 0.751 (c) Since n1 n2 , d.f. = n 1 = 30 1 = 29. From Table 6 in Appendix II, 0.751 falls between entries 0.683 and 1.174. Use two-tailed areas to find 0.250 < P value < 0.500. Using a TI-84, d.f. ≈ 57.92, and P value ≈ 0.4556. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 529 (d) Since the P-value interval > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to indicate there is a difference between the control and experimental groups in the mean score on the vocabulary portion of the test. 16. (a) 0.01 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. Both sample sizes are large, ni 30, and 1 and 2 are unknown. x1 x2 368.4 349.2 19.2 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n 2 19.2 0 39.52 30 56.62 30 1.524 (c) Since n1 n2 , d.f. = n 1 = 30 1 = 29. From Table 6 in Appendix II, 1.524 falls between entries 1.479 and 1.699. Use one-tailed areas to find 0.050 < P value < 0.075. Using a TI-84, d.f. ≈ 51.83, and P value ≈ 0.0668. (d) Since the P-value interval > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, the evidence is insufficient to claim that the population mean score for the experimental group was higher than for the control group. 17. (i) Use a calculator to verify. Use rounded results to compute t. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 530 (ii) (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume that the population distributions for both are approximately normal and mound-shaped and that 1 and 2 are unknown. x1 x2 4.75 3.93 0.82 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n2 0.82 0 2.822 16 2 0.869 2.43 15 (c) Since n2 n1 , d . f . n2 1 15 1 14. From Table 6 in Appendix II, 0.869 falls between entries 0.692 and 1.200. Use two-tailed areas to find 0.250 < P value < 0.500. Using a TI-84, d.f. ≈ 28.81, and P value ≈ 0.1961. (d) Since the P-value interval is > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to indicate that there is a difference in the mean number of cases of fox rabies between the two regions. 18. (i) Use a calculator to verify. Use rounded results to compute t. (ii) (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume the population distributions for both are approximately normal and mound-shaped and that 1 and 2 are unknown. x1 x2 12.53 10.77 1.76 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n 2 1.76 0 2.392 14 2.402 16 2.008 (c) Since n1 n2 , d . f . n1 1 14 1 13. From Table 6 in Appendix II, 2.008 falls between entries 1.771 and 2.160. Use one-tailed areas to find 0.025 < P value < 0.050. Using a TI-84, d.f. ≈ 27.50, and P value ≈ 0.0273. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 531 (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to indicate that population mean soil water content is higher in Field A. 19. (i) Use a calculator to verify. Use rounded results to compute t. (ii) (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume the population distributions for both are approximately normal and mound-shaped and that 1 and 2 are unknown. x1 x2 4.86 6.5 1.64 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n 2 1.64 0 3.182 7 2.882 8 1.041 (c) Since n1 n2 , d . f . n1 1 7 1 6. From Table 6 in Appendix II, 1.041 falls between entries 0.718 and 1.273. Use two-tailed areas to find 0.250 < P value < 0.500. Using a TI-84, d.f. ≈ 12.28, and P value ≈ 0.3179. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 532 (d) Since the P-value interval is > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to indicate that the mean time lost owing to hot tempers is different from that lost owing to technical workers’ attitudes. 20. (a) d 2.25; sd 7.778; t0.95 = 2.365 with d.f. = 7; E 2.365; 95% confidence interval for d from – 4.25 to 8.75. We are 95% confident that the difference in population mean percentage increase between company revenue and CEO salaries is between –4.25% and 8.75%. (b) = 0.05; c = 1 – = 0.95; H0: d = 0; H1: d ≠ 0; since d = 0 from the null hypothesis is in the 95% confidence interval, do not reject H0 at the 5% level of significance. The data do not indicate a difference in population mean percentage increases between company revenue and CEO salaries. This result is consistent with the conclusion reached using the P-value method of testing for = 0.05 in Problem 7. 21. (a) d . f . s12 s22 n1 n2 2 2 0.812 0.942 10 12 2 2 1 0.812 1 0.942 11 9 10 12 19.96 2 2 s22 1 n2 1 n2 (Note: Some software will truncate this to 19.) (b) In Problem 13, d . f . n1 1 10 1 9. The convention of using the smaller of n1 1 and n2 1 leads to a d.f. that is always less than or equal to that computed by Satterthwaite’s formula. 1 s1 n1 1 n1 2 (n1 1) s12 (n2 1) s22 22. (a) s n1 n2 2 x1 x2 t 15(2.822 ) 14(2.432 ) 2.639 16 15 2 4.75 3.93 0.865 1 1 2.639 16 15 d . f . n1 n2 2 16 15 2 29 H0 : 1 2 H1: 1 2 From Table 6 in Appendix II, 0.865 falls between entries 0.683 and 1.174. Using two-tailed areas, find 0.250 < P value < 0.500. Since the P-value interval > 0.05, we fail to reject H 0 . (b) Using unpooled sample standard deviation, t 0.869, d.f. = 14, and fail to reject H 0 . Conclusions are the same, however; the exact P value using the pooled standard deviation method and larger d.f. will be slightly smaller than that using unpooled methods. s 1 n1 1 n2 23. (a) 0.05 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 533 r1 r2 59 56 0.2911 n1 n2 220 175 q 1 p 1 0.2911 0.7089 p r1 r 59 56 0.268, pˆ 2 2 0.32 n1 220 n2 175 pˆ1 pˆ 2 0.268 0.32 0.052 pˆ pˆ 2 0.052 z 1 1.13 pq pq 0.2911(0.7089) 0.2911(0.7089) n n 220 175 pˆ1 1 2 (c) P value = 2P(z < 1.13) = 2(0.1292) = 0.2584 (d) Since the P value = 0.2584 > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, there is insufficient evidence to conclude that the population proportion of women favoring more tax dollars for the arts is different from the proportion of men. 24. (a) 0.05 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. r1 r2 21 22 0.2443 n1 n2 93 83 q 1 p 1 0.2443 0.7557 p r1 21 r 22 0.226, pˆ 2 2 0.265 n1 93 n2 83 pˆ1 pˆ 2 0.226 0.265 0.039 pˆ pˆ 2 0.039 z 1 0.61 pq pq 0.2443(0.7557) 0.2443(0.7557) n n 93 83 pˆ1 1 2 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 534 (c) P value = P(z < 0.61) = 0.2709 (d) Since the P value = 0.2709 > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, there is insufficient evidence to conclude that the population proportion of conservative voters favoring more tax dollars for the arts is less than the proportion of moderate voters. 25. (a) 0.01 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. r1 r2 12 7 0.0676 n1 n2 153 128 q 1 p 1 0.0676 0.9324 p r1 12 r 7 0.0784, pˆ 2 2 0.0547 n1 153 n2 128 pˆ1 pˆ 2 0.0784 0.0547 0.0237 pˆ pˆ 2 0.0237 z 1 0.79 pq pq 0.0676(0.9324) 0.0676(0.9324) n n 153 128 pˆ1 1 2 (c) P value = 2P(z > 0.79) = 2(0.2148) = 0.4296 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 535 (d) Since the P value = 0.4296 > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, there is insufficient evidence to conclude that the population proportion of high-school dropouts on Oahu is different from that of Sweetwater County. 26. (a) 0.05 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. r1 r2 141 125 0.5278 n1 n2 288 216 q 1 p 1 0.5278 0.4722 p r1 141 r 125 0.4896, pˆ 2 2 0.5787 n1 288 n2 216 pˆ1 pˆ 2 0.4896 0.5787 0.0891 pˆ pˆ 2 0.0891 z 1 1.98 pq pq 0.5278(0.4722) 0.5278(0.4722) n n 288 216 pˆ1 1 2 (c) P value = P(z < 1.98) = 0.0239 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 536 (d) Since the P value = 0.0239 < 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, there is sufficient evidence to conclude that the population proportion of voter turnout in Colorado is greater than that in California. 27. (a) 0.01 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. r1 r2 37 47 0.42 n1 n2 100 100 q 1 p 1 0.42 0.58 p r1 r 37 47 0.37, pˆ 2 2 0.47 n1 100 n2 100 pˆ1 pˆ 2 0.37 0.47 0.10 pˆ pˆ 2 0.10 z 1 1.43 pq pq 0.42(0.58) 0.42(0.58) n n 100 100 pˆ1 1 2 (c) P value = P(z < -1.43) 0.0764 (d) Since the P value = 0.0764 > 0.01, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 1% level of significance, there is insufficient evidence to conclude that the population proportion of adults believing in extraterrestrials who attended college is higher than the proportion who did not attend college. 28. (a) 0.05 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 537 r1 r2 45 36 0.6694 n1 n2 59 62 q 1 p 1 0.6694 0.3306 p r1 45 r 36 0.7627, pˆ 2 2 0.5806 n1 59 n2 62 pˆ1 pˆ 2 0.7627 0.5806 0.182 pˆ pˆ 2 0.182 z 1 2.13 pq pq 0.6694(0.3306) 0.6694(0.3306) n n 59 62 pˆ1 1 2 (c) P value = P(z > 2.13) = 0.0166 (d) Since the P value = 0.0166 < 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, there is sufficient evidence to conclude that the population proportion of conservative voters who prefer art with fully clothed people is higher. 29. (a) 0.05 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. r1 r2 45 71 0.2189 n1 n2 250 280 q 1 p 1 0.2189 0.7811 p r1 r 45 71 0.18, pˆ 2 2 0.2536 n1 250 n2 280 pˆ1 pˆ 2 0.18 0.2536 0.074 pˆ pˆ 2 0.074 z 1 2.04 pq pq 0.2189(0.7811) 0.2189(0.7811) n n 250 280 pˆ1 1 2 (c) P value = P(z < 2.04) = 0.0207 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 538 (d) Since the P value = 0.0207 < 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, there is sufficient evidence to conclude that the population proportion of trusting people in Chicago is higher in the older group. 30. H0 : 1 2 ; H1: 1 2 ; for = 0.01 and a right-tailed test, the critical value z0 2.33; sample test statistic z = 2.57 is in the critical region, reject H 0 . This result is consistent with the P-value method. 31. H0 : 1 2 ; H1: 1 2 ; for d.f. = 9, = 0.01 in the one-tail area row, the critical value t0 2.821. Sample test statistic t = 0.965 is not in the critical region, fail to reject H 0 . This result is consistent with the P-value method. 32. H 0 : p1 p2 ; H1: p1 p2 ; for = 0.05 and a two-tailed test, the critical values are z0 1.96; sample test statistic z = 1.13 is not in the critical region, fail to reject H 0 . This result is consistent with the Pvalue method. Chapter 9 Review 1. Look at the original x distribution. If it is normal or n ≥ 30, and σ is known, use the standard normal distribution. If the x distribution is mound-shaped or n ≥ 30, and σ is unknown, use the Student’s t distribution. The degrees of freedom are determined by the application. 2. A significant test means we reject H0. Statistically significant results are not necessarily practically significant results. 3. A larger sample size increases the value of z or t. 4. A larger value of z or t will have a smaller P value. 5. (a) 0.05 H 0 : 11.1 H1: 11.1 Since is in H1 , use a two-tailed test. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 539 (b) Use the standard normal distribution. We assume that x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x n 10.8 11.1 0.6 36 3.00 (Note: 600 miles = 0.6 thousand miles) (c) P value = 2P(z < 3.00) = 2(0.0013) = 0.0026 (d) Since a P value of 0.0026 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to say that the miles driven per vehicle in Chicago is different from the national average. 6. (a) 0.05 H 0 : p 0.35 H1: p 0.35 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 81(0.35) = 28.35 and nq = 81(0.65) = 52.65, and both are greater than 5. r 39 pˆ 0.4815 n 81 pˆ p 0.4815 0.35 z 2.48 pq n 0.35(0.65) 81 (c) P value = P(z > 2.48) = 0.0066 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 540 (d) Since a P value of 0.0066 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence indicates that more than 35% of the students have jobs. 7. (a) 0.01 H 0 : 0.8 A H1: 0.8 A (b) Use the Student’s t distribution with d.f. = n 1 = 9 1 = 8. We assume that x has a distribution that is approximately normal and that is unknown. x 1.4 0.8 t 4.390 s n 0.41 9 (c) For d.f. = 8, 4.390 falls between entries 3.355 and 5.041. Using one-tailed areas, 0.0005 < P value < 0.005. Using a TI-84, P value ≈ 0.0012. Student's t Distribution, d.f. = 8 -3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 4.39 5.00 (d) Since the entire P-value interval 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to say that the Toylot claim of 0.8 A is too low. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 8. 541 (a) 0.01 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume that both population distributions are approximately normal and that 1 and 2 are unknown. x1 x2 9.4 6.9 2.5 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n2 2.5 2.12 12 2 2.986 2.0 12 (c) Since n1 n2 , d.f. = n 1 = 12 1 = 11. In Table 6 in Appendix II, 2.986 falls between entries 2.718 and 3.106. Use one-tailed areas to find that 0.005 < P value < 0.010. Using a TI-84, d.f. ≈ 21.95, and P value ≈ 0.0034. (d) Since the P-value interval is 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the evidence shows that the yellow paint has less visibility after 1 year. 9. (a) 0.01 H 0 : p 0.60 H1: p 0.60 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 90(0.6) = 54 and nq = 90(0.4) = 36, and both are greater than 5. r 40 pˆ 0.4444 n 90 pˆ p 0.4444 0.60 z 3.01 pq n 0.60(0.40) 90 (c) P value = P(z < 3.01) = 0.0013 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 542 (d) Since a P value of 0.0013 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the evidence is sufficient to show the mortality rate has dropped. 10. (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume that both population distributions are approximately normal and that 1 and 2 are unknown. x1 x2 53 62 9 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n2 9 0 192 51 2 2.735 15 60 (c) Since n1 n2 , d . f . n1 1 51 1 50. In Table 6 in Appendix II, 2.735 falls between entries 2.678 and 3.496. Use two-tailed areas to find that 0.001 < P value < 0.010. Using a TI-84, d.f. ≈ 94.53, and P value ≈ 0.0074. (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence shows that there is a significant difference in average offschedule times. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 543 11. (a) 0.01 H 0 : 40 H1: 40 Since > is in H1 , use a right-tailed test. (b) Use the standard normal distribution. We assume that x has a normal distribution with known standard deviation . Note that is given in the null hypothesis. z x 43.1 40 9 94 n 3.34 (c) P value = P(z > 3.34) = 0.0004 Standard Normal Distribution -4.00 -3.00 -2.00 -1.00 0.00 1.00 2.00 3.34 4.00 (d) Since a P value of 0.0004 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, the evidence is sufficient to say that the population average number of matches is larger than 40. 12. (a) 0.05 H 0 : p1 p2 H1: p1 p2 (b) Use the standard normal distribution. The number of trials is sufficiently large because n1 p, n1q , n2 p, and n2 q are each larger than 5. r1 r2 12 18 0.1622 n1 n2 88 97 q 1 p 1 0.1622 0.8378 p r1 12 r 18 0.1364, pˆ 2 2 0.1856 n1 88 n2 97 pˆ1 pˆ 2 0.1364 0.1856 0.0492 pˆ pˆ 2 0.0492 z 1 0.91 pq pq 0.1622(0.8378) 0.1622(0.8378) n n 88 97 pˆ1 1 2 (c) P value = P(z < 0.91) = 0.1814 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 544 (d) Since the P value = 0.1814 > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, there is insufficient evidence to say that the population proportion of suburban residents subscribing to Sporting News is higher. 13. (a) 0.05 H 0 : 1 2 H1: 1 2 (b) Use the Student’s t distribution. We assume that both population distributions are approximately normal and that 1 and 2 are unknown. x1 x2 3.0 2.7 0.3 t ( x1 x2 ) ( 1 2 ) s12 n1 s22 n2 0.3 0 0.82 55 2 1.808 0.9 51 (c) Since n2 n1 , d . f . n2 1 51 1 50. In Table 6 in Appendix II, 1.808 falls between entries 1.676 and 2.009. Use two-tailed areas to find that 0.050 < P value < 0.100. Using a TI-84, d.f. ≈ 100.27, and P value ≈ 0.0735. (d) Since the P-value interval > 0.05, we fail to reject H 0 . No, the data are not statistically significant. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 545 (e) At the 5% level of significance, the evidence is insufficient to show that there is a difference in population mean length of the two types of projectile points. 14. (a) 0.05 H 0 : p 0.36 H1: p 0.36 (b) Use the standard normal distribution. The p̂ distribution is approximately normal when n is sufficiently large, which it is here, because np = 120(0.36) = 43.2 and nq = 120(0.64) = 76.8, and both are greater than 5. r 33 pˆ 0.275 n 120 pˆ p 0.275 0.36 z 1.94 pq n 0.36(0.64) 120 (c) P value = P(z < 1.94) = 0.0262 (d) Since a P value of 0.0262 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, the evidence is sufficient to show that the population percentage of bachelor or higher degree holders in the private sector is less than in the federal civilian sector. 15. (a) 0.05 H 0 : 7 oz H1: 7 oz (b) Use the Student’s t distribution with d.f. = n 1 = 8 1 = 7. We assume that x has a distribution that is approximately normal and that is unknown. x 7.3 7 t 1.697 s n 0.5 8 (c) For d.f. = 7, 1.697 falls between entries 1.617 and 1.895. Using two-tailed areas, 0.100 < P value < 0.150. Using a TI-84, P value ≈ 0.1335. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 546 (d) Since the entire P-value interval > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, the evidence is insufficient to show that the population mean amount of coffee per cup is different from 7 ounces. 16. (a) 0.01 H 0 : d 0 H1: d 0 (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d = exp cont A 14 B 14 C 7 D 7 E 12 F 5 d 9.83, sd 3.97 t d d 9.83 0 6.07 sd n 3.97 6 (c) d.f. = n 1 = 6 1 = 5 From Table 6 in Appendix II, 6.07 falls between entries 4.032 and 6.869. Using one-tailed areas, find 0.0005 < P value < 0.005. Using a TI-84, P value ≈ 0.0009. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 547 Student's t Distribution, d.f. = 5 -3.000 -2.000 -1.000 0.000 1.000 2.000 3.000 4.000 5.000 6.066 7.000 (d) Since the P-value interval is 0.01, we reject H 0 . Yes, the data are statistically significant. (e) At the 1% level of significance, there is sufficient evidence to indicate that the program of the experimental group promoted creative problem solving. 17. (a) 0.05 H 0 : d 0 H1: d 0 (b) Use the Student’s t distribution. Assume that d has a normal distribution or has a mound-shaped, symmetric distribution. Pair d = before after 1 6.4 2 7.2 3 8.6 4 3.8 5 1.3 d 4.94, sd 3.90 t d d sd n 4.94 0 3.90 5 2.832 (c) d.f. = n 1 = 5 1 = 4 From Table 6 in Appendix II, 2.832 falls between entries 2.776 and 3.747. Use one-tailed areas to find that 0.010 < P value < 0.025. Using a TI-84, P value ≈ 0.0236. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 548 (d) Since the P-value interval is 0.05, we reject H 0 . Yes, the data are statistically significant. (e) At the 5% level of significance, there is insufficient evidence to claim that the injection system lasts less than an average of 48 months. 18. (i) Use a calculator to verify. The rounded values are used in part (ii). (ii) (a) 0.05 H 0 : 48 months H1: 48 months (b) Use the Student’s t distribution with d.f. = n 1 = 10 1 = 9. We assume that x has a distribution that is approximately normal and that is unknown. x 44.2 48 t 1.396 s n 8.61 10 (c) For d.f. = 9, 1.396 falls between entries 1.383 and 1.574. Use one-tailed areas to find that 0.075 < P value < 0.100. Using a TI-84, P value ≈ 0.0981. (d) Since the entire P-value interval > 0.05, we fail to reject H 0 . No, the data are not statistically significant. (e) At the 5% level of significance, there is insufficient evidence to claim that the injection system lasts less than an average of 48 months. Cumulative Review Chapters 7–9 1. (a) Answers vary. A population is the entire group of interest. Random sampling is accomplished through the assigning of numbers to members of the population and then by using a random-number table or a computer package to draw the simple random sample. (b) A sample statistic is a numerical value calculated from the sample. A sampling distribution is the probability distribution associated with a sample statistics such as the sample mean. (c) See Chapter 7. (d) Answers vary. 2. (a) Since the sample size is large, the central limit theorem describes the distribution of x . (b) P( x 6820) P( z 2.75) 0.0030 z 6820 7500 2.75 1750 50 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 549 (c) The probability that the 50 workers’ average white blood cell count is that low or lower, if they were healthy adults, is extremely low. It would be reasonable to gather additional facts. 3. We are given p = 0.45 and n = 32. Both np > 5 and nq > 5. (a) 18 0.5 pˆ 0.5625 Continuity correction 0.0156 32 32 P( pˆ 0.5625) P(0.5625 pˆ ) P(0.5625 0.0156 x) P(0.5469 x) P(1.10 z ) 0.1357 0.5469 0.45 z 1.10 0.45 0.55 32 (b) 10 0.5 0.3125 Continuity correction 0.0156 32 32 P( pˆ 0.3125) P( pˆ 0.3125 0.0156) pˆ P( x 0.3281) P( z 1.39) 0.0823 0.3281 0.45 z 1.39 0.45 0.55 32 (c) Yes. Here, np = 32 × 0.45 = 14.4 and nq = 32 × 0.55 = 17.6, both greater than 5. The expected value of pˆ is p 0.45 . The standard deviation of pˆ is pˆ 4. pq n 0.45 0.55 0.088 . 32 (a) H0: μ = 2.0 H1: μ > 2.0 α = 0.01 Since σ is known and x is normal, we will use the standard normal distribution. 2.56 2.0 z 2.53 0.7 10 The P value will be P(z > 2.53) = 0.0057. Since the P value = 0.0057 < 0.01, we reject H0. Yes, the data are statistically significant. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 550 At the 1% level of significance, the evidence is sufficient to say that the population mean discharge level of lead is higher. (b) x z n x z 2.56 1.96 0.7 10 n 2.56 1.96 0.7 10 2.13 2.99 2 2 z 1.96 0.7 (c) n c 47.06, so use n = 48 water samples. E 0.2 5. (a) Use a calculator to confirm that x 10.49 and s ≈ 1.36. (b) H0: μ = 10% H1: μ > 10% α = 0.05 Since x is approximately normal and σ is unknown, we will use a Student’s t distribution. d.f. = 12 – 1 = 11 10.49 10 t 1.248 1.36 12 Using a TI-84, the P value will be P(t > 1.248) = 0.119. Using Table 6 in Appendix II, 0.100 < P value < 0.125. Since the P-value interval > 0.05, we fail to reject H0. The data are not statistically significant. At the 5% level of significance, the evidence does not indicate that the patient is asymptotic. (c) For a 99% interval, tc = 3.106. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 x t s n x t s 10.49 3.106 1.36 12 551 n 10.49 3.106 1.36 12 9.27% 11.71% 6. (a) Check conditions: np 68 0.10 6.8, nq 68 0.90 61.2. Since both estimates are reasonably greater than 5, the conditions are satisfied. H0: p = 0.10 H1: p ≠ 0.10 α = 0.05 Since the conditions are satisfied, p̂ follows a normal distribution. z 0.147 0.10 1.29 0.10 0.90 68 The P value will be 2 × P(z > 1.29) = 0.197. Since the P value > 0.05, we fail to reject H0. No, the data are not statistically significant. At the 5% level of significance, the data do not indicate any difference from the national average for the population proportion of crime victims. (b) For a 95% confidence interval, zc = 1.96. ˆˆ ˆˆ pq pq pˆ z p pˆ z n n 0.147 1.96 0.147 0.853 0.147 0.853 p 0.147 1.96 68 68 0.063 p 0.231 2 1.96 (c) n 0.147 0.853 192.68, so use n = 193 students. 0.05 Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 552 7. H0: μd = 0 H1: μd ≠ 0 α = 0.05 Pair d=A–B 1 –0.001 2 –0.028 3 –0.002 4 0.016 5 –0.01 6 0.001 7 –0.003 Here, d 0.0039 , sd = 0.013, and we will use the Student’s t distribution with d.f. = 7 – 1 = 6. 0.0039 0 t 0.771 0.013 7 Using a TI-84, the P value will be 2 × P(t < –0.771) = 0.470. Using Table 6 in Appendix II, 0.250 < P value < 0.500. Since the P-value interval > 0.05, we fail to reject H0. No, the data are not statistically significant. At the 5% level of significance, the evidence does not show a difference in phosphorous reduction between the two methods. 8. (a) H0: μ1 = μ2 H 1: μ 1 ≠ μ 2 α = 0.05 We will use the Student’s t distribution with d.f. = 16 – 1 = 15. 74.8 70.1 0 0 t 1.952 5.22 8.62 16 18 Using the TI-84, the P value will be 2 × P(t > 1.952) = 0.070. Using Table 6 in Appendix II, 0.050 < P value < 0.100. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 553 Since the P-value interval > 0.05, we fail to reject H0. No, the data are not statistically significant. At the 5% level of significance, the evidence does not show any difference in the population mean proportion of on-time arrivals in summer versus winter. (b) x1 x2 t s12 s2 2 s2 s 2 1 2 x1 x2 t 1 2 n1 n2 n1 n2 74.8 70.1 2.131 5.22 8.62 5.22 8.62 1 2 74.8 70.1 2.131 16 18 16 18 0.43% 1 2 9.83% (c) We assume that x1 and x2 are approximately normal or at least mound-shaped and symmetric. 9. (a) H0: p1 = p2 H1: p1 > p2 α = 0.05 We check that n1 pˆ1 23, n2 pˆ 2 19, n1qˆ1 72, and n2 qˆ2 73, and all are greater than 5. It thus seems reasonable to use the normal distribution. 23 19 p 0.2246 95 92 0.242 0.207 0 0 z 0.58 0.2246 0.7754 0.2246 0.7754 95 92 The P value will be P(z > 0.58) = 0.2810. Copyright © Houghton Mifflin Company. All rights reserved. Part IV: Complete Solutions, Chapter 9 554 Since P value > 0.05, we fail to reject H0. No, the data are not statistically significant. At the 5% level of significance, the evidence does not indicate that the population proportion of single men is greater than the population proportion of single women. (b) pˆ1 pˆ 2 z pq pq p1 p2 pˆ1 pˆ 2 z n1 n2 pq pq n1 n2 0.242 0.207 1.645 0.2246 0.7754 0.2246 0.7754 p1 p2 95 92 0.242 0.207 1.645 0.2246 0.7754 0.2246 0.7754 95 92 0.065 p1 p2 0.139 10. (a) Essay. Refer to Chapters 7, 8, and 9. (b) Answers will vary. 11. Answers will vary. Salient points include that the hypotheses of interest are H0: μ1 = μ2 H1: μ1 < μ2, with μ1 representing the average time to recovery under the new method At the 1% level of significance, the data are statistically significant, indicating that the new method leads to shorter recovery times, on average. Copyright © Houghton Mifflin Company. All rights reserved.